Under current regulatory policies, chemicals that are believed to comprise both a significant exposure and health risk in the human population are selected to undergo testing for toxicity and carcinogenic potency. The traditional method for evaluating carcinogenic activity and chronic toxicity of a specific chemical has been the two-year bioassay. Due to relatively large amounts of resources involved, each bioassay costs between 2 million and 4 million dollars and takes several years to complete. According to the National Toxicology Program (NTP), the number of chemicals currently tested stands at 505 in long-term studies and 66 in short-term tests, and only a single sub-chronic study. By comparing the number of completed bioassays with the 70,000 to 85,000 chemicals in commerce today it is impossible to apply the current testing system to all chemicals of concern. Alternative approaches must be developed in the interest of protecting human health and in saving economic resources. To develop an efficient and accurate assessment of the carcinogenic potential of an unknown chemical we propose to: (1) identify a set of genes that is predictive of chemically induced hepatocarcinogenesis in a standard rodent 2-year bioassay using a tissue-engineered rodent liver model; and (2) apply the diagnostic gene expression profile derived from the rodent studies to the equivalent tissue-engineered human liver model to evaluate cross-species scaling of the chemically-induced carcinogenic endpoint. To accomplish this, in vivo exposures and exposures of engineered three-dimensional liver co-cultures will be performed with a training set of compounds that include non-hepatocarcinogens and hepatocarcinogens of the following classes: genotoxic, non-genotoxic, sex-dependent and species-dependent. Microarray analysis will be performed on mRNA derived from the exposed animal and liver cultures and the results used to identify gene sets that are statistically predictive of hepatocarcinogenicity in the two-year animal bioassay. A commercial product derived from this research will enable prioritization of hepatocarcinogenic potential of untested chemicals and allow a significant savings of time, money and animals for both governmental agencies as well as private industry related to chemical manufacturing, food additives and pharmaceuticals.